1. Field of the Invention
The present invention generally relates to a self-aligned double-gate metal oxide semiconductor (DG-MOSFET), with electrically separated top and bottom gates. Moreover, with the invention, the top and bottom gates may be formed by different materials.
2. Description of the Related Art
The double-gate metal oxide semiconductor field effect transistor (DG-MOSFET), is a MOSFET having a top and a bottom gate which control the carriers in the channel. The double-gate MOSFET has several advantages over a conventional single-gate MOSFET: higher transconductance, lower parasitic capacitance, avoidance of dopant fluctuation effects, and superior short-channel characteristics. Moreover, good short-channel characteristics are obtained down to 20 nm channel length with no doping needed in the channel region. This circumvents all the tunneling break-down, dopant quantization, and impurity scattering problems associated with channel doping.
Conventional systems have attempted to make a double-gate structure with both top and bottom gates self-aligned to the channel region. However, there is no satisfactory method of achieving this self-aligned structure. Previous efforts generally fall into the following categories. A first, category includes etching silicon (Si) into a pillar structure and depositing gates around it (vertical Field Effect Transistor (FET)). A second, category etches a silicon on insulator (SOI) film into a thin bar, makes the source/drain contacts on both ends of the bar, and deposits the gate material on all three surfaces of the thin Si bar. Another way involves making a conventional single-gate MOSFET, then using bond-and-etch back techniques to form the second gate. A fourth conventional method starts with a thin SOI film, patterns a strip and digs a tunnel under it by etching the buried oxide to form a suspended Si bridge. Then, this method deposits the gate material all around the suspended Si bridge.
There are serious drawbacks in all of the above approaches. For example, the first and second require formation of a vertical pillar or Si bar at a thickness of 10 nm and it is difficult to reach this dimension with good thickness control and prevent Reactive Ion Etching (RIE) damage. While in the vertical case (first), it is difficult to make a low series resistance contact to the source/drain terminal which is buried under the pillar. In the lateral case (second), the device width is limited by the Si bar height. In the third case, thickness control and top/bottom gate self-alignment are major problems. In the fourth case, the control over the gate length is poor, and the two gates are electrically connected and must be made of the same material.
A co-pending application by, K. K. Chan, G. M. Cohen, Y. Taur, H. S. P. Wong, entitle “Self-Aligned Double-Gate MOSFET by Selective Epitaxy and Silicon Wafer Bonding Techniques”, Ser. No. 09/272,297, filed Mar. 19, 1999 (hereinafter “Chan”) incorporated herein by reference, utilizes a method for the fabrication of a double-gate MOSFET structure with both top and bottom gates self-aligned to the channel region. The process circumvents most of the problems discussed above. Yet, the top and bottom gates are still physically connected. This occurs because the gate material is deposited in one processing step as an “all-around the channel” gate.
This may not be desirable in some applications for the following reasons. First, from the circuit design point of view, two electrically separated gates are preferable. Second, the bottom gate and top gate are essentially made of the same material, thus only a symmetric DG-MOSFET may be fabricated. Asymmetric DG-MOSFET in which the bottom gate material is different than the top gate cannot be realized.
Chan discloses forming an “all-around the channel” gate by forming a suspended silicon bridge (the channel) followed by the deposition of the gate material conformally around it. To obtain a good threshold voltage control, the channel thickness should be thinned down to 3-5 nm. It is not clear if such thin bridges can be processed with a high enough yield. Thus, this may impose a limitation on the process suggested in Chan.
Thus, there is a need for a self-aligning DG-MOSFET that is formed by depositing the top and bottom gates independently. Such a structure would produce many advantages. For example, the independent formation of the gates permits the gates to be electrically separated; to be made of varying materials and thickness, and to provide a structure that is planarized, making it easier to connect the device. In addition, there is a need for a DG-MOSFET which permits the formation of a very thin channel.